|
From: | John W. Eaton |
Subject: | Re: prod auto-conversion to double |
Date: | Mon, 21 Oct 2013 21:14:23 -0400 |
User-agent: | Mozilla/5.0 (X11; Linux x86_64; rv:10.0.12) Gecko/20130116 Icedove/10.0.12 |
On 10/21/2013 05:59 PM, Rik wrote:
10/21/13 All, I was investigating factor() for large numbers above bitmax(), but still below intmax(), and came across this behavior. Is this desirable? Should we have a "native" option as we do for sum() so that the full precision can be realized? --Rik Sample case: x = int64 ([3, 107, 28059810762433]); class (x) ans = int64 y = prod (x) y = 9007199254740992 class (y) ans = double x(1) * x(2) * x(3) ans = 9007199254740993
Yes, it looks like this is already something that Matlab has for prod as well as sum.
jwe
[Prev in Thread] | Current Thread | [Next in Thread] |